Multivariate patterns of biochemical responses of Pinus ponderosa trees atfield plots in the San Bernardino Mountains, southern California

Most environmental stress conditions promote the production of potentially toxic active oxygen species in plant cells. Plants respond with changes in their antioxidant and photoprotective systems. Antioxidants and pigments ha ve been widely used to measure these responses. Because trees are exposed t o multiple man-made and natural stresses, their responses are not reflected by changes in single stress markers, but by complex biochemical changes. T o evaluate such response patterns, explorative multivariate statistics have been used. In the present study, 12 biochemical variables (chloroplast pig ments, state of the xanthophyll cycle, alpha -tocopherol, ascorbate and deh ydroascorbate, glutathione and oxidized glutathione) were measured in previ ous-year needles of field-grown Pinus ponderosa Dougl. ex Laws. The trees w ere sampled in two consecutive years in the San Bernardino Mountains in sou thern California, where a pollution gradient is overlaid by gradients in na tural stresses (drought, altitude). To explore irradiance effects, needle s amples were taken directly in the field (sun exposed) and from detached, da rk-adapted branches. A principal component analysis on this data set (n = 8 0) resulted in four components (Components 1-4) that explained 67% of the v ariance in the original data. Component 1 was positively loaded by concentr ations of alpha -tocopherol, total ascorbate and xanthophyll cycle pools, a s well as by the proportion of de-epoxides in the xanthophyll cycle. It was negatively loaded by the proportion of dehydroascorbate in the ascorbate p ool. Component 2 was negatively loaded by chlorophyll concentrations, and p ositively loaded by the ratios of lutein and beta -carotene to chlorophyll and by the de-epoxidation state of the xanthophyll cycle. Component 3 was n egatively loaded by GSH concentrations and positively loaded by the proport ions of GSSG and tocopherol concentrations. Component 4 was positively load ed by neoxanthin and negatively loaded by beta -carotene. The four componen ts could be assigned to the concerted action of the biochemical protection system: high scores on Component 1 represent highly activated antioxidative defense, changes in pigment composition are represented in Components 2 an d 4, and the glutathione system, which is important for antioxidant regener ation, is represented in Component 2. Although Component 1 scores were gene rally higher (indicating activation of antioxidant defense) in light-adapte d needles relative to dark-adapted needles, they were also site dependent w ith increased scores at sites with less pollution, but higher natural stres s impacts. High scores of Components 2 and 3 at the highest elevation site, which was only moderately polluted, indicated an increase in photoprotecti on by pigments and activation of the glutathione system. Significant differ ences between light- and dark-adapted needles in Components 2 and 3 were on ly found at the site with the highest pollution. Use of accumulated variabl es (components) instead of single biochemical variables enabled recognition of response patterns at particular sites and a better comparison with resu lts of other studies is expected. Typical response patterns could be assign ed to particular environmental stress combinations, providing a means of as sessing potential biological risks within individual forest stands.